Research On Medical Image Visualization Technology Based On WEB

The traditional medical imaging visualization systems are based on a specific platform or visual plug-in.It is difficult to achieve cross-platform and remote access.In the process of using,it increases the cost and difficulty of post-system maintenance,which also largely restricts the development of the medical image visualization system.However,as the WEBGL function becomes more and more powerful,it can easily create high-performance computer graphics by calling the Graphics card computing resources.This makes it possible affords to visualize medical images on the browser.Therefore,this paper proposes and implements a Web-oriented medical image visualization system based on HTML5 and WebGL.The design of system architecture is based on the B/S model.This system provides DICOM format medical image visualization services.The system uses WebGL to accelerate the hardware,draw real-time two-dimensional medical images on the web and improve the speed and quality of 3D reconstruction by using improved ray casting algorithm.This paper solves the problem of slow 3D reconstruction by improving the traditional light projection algorithm based on WebGL.From the selection of sampling points,interpolation operations,early light cut-off to improve the algorithm.In this paper,we use the acceleration step sampling algorithm to determine the position of sampling points,and utilize the improved interpolation algorithm to calculate the resample points,which can effectively improve the computing speed of sampling and image synthesis and accelerate the 3D reconstruction speed while ensuring the display quality.Finally,the light projection cut-off method was used to further improve the speed of 3D reconstruction.The above improved algorithm is implemented in the fragment shader through WebGL.Compared with the traditional light projection algorithm,this improved algorithm plays a certain accelerating effect,drawing speed is obviously improved.The experimental results show that this system is tested for different browsers and multiple groups of clinical DICOM images.The system can be very friendly to support cross-browser visualization,showing good results and fast speed.The system supports two-dimensional real-time visualization of DICOM medical image data as well as two-dimensional image anti-color displayed,pseudo-color,zoom,pan and other interactive operations.At the same time,3D reconstruction based on improved light projection also has a good visualization effect,and various interactive functions can run smoothly.